File set-up

Set working directory to current directory

if (rstudioapi::isAvailable()) {
  setwd(dirname(rstudioapi::getActiveDocumentContext()$path))
}

Load standard libraries and resolve conflicts

library(tidyverse)

## ── Attaching packages ─────────────────────────────────────── tidyverse 1.3.2 ──
## ✔ ggplot2 3.3.6      ✔ purrr   0.3.4 
## ✔ tibble  3.1.8      ✔ dplyr   1.0.10
## ✔ tidyr   1.2.0      ✔ stringr 1.4.1 
## ✔ readr   2.1.2      ✔ forcats 0.5.2 
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag()    masks stats::lag()

library(conflicted)
conflict_prefer("filter", "dplyr")

## [conflicted] Will prefer dplyr::filter over any other package

conflict_prefer("select", "dplyr")

## [conflicted] Will prefer dplyr::select over any other package

conflict_prefer("slice", "dplyr")

## [conflicted] Will prefer dplyr::slice over any other package

conflict_prefer("rename", "dplyr")

## [conflicted] Will prefer dplyr::rename over any other package

conflict_prefer('intersect', 'dplyr')

## [conflicted] Will prefer dplyr::intersect over any other package

Read data

cq = read_tsv('../data/Supplementary_Table_3_selected_circRNAs.txt')

## Rows: 1560 Columns: 44
## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: "\t"
## chr (24): chr, strand, circ_id, circ_id_strand, tool, cell_line, FWD_primer,...
## dbl (20): start, end, FWD_pos, FWD_length, REV_pos, REV_length, FWD_TM, REV_...
## 
## ℹ Use `spec()` to retrieve the full column specification for this data.
## ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

cq

Presence in db (n = 890)

make count table

cont_table = cq %>%
  select(circ_id, cell_line, qPCR_val, RR_val, amp_seq_val, count_group, n_db) %>%
  unique() %>%
  # only keep high abundant circRNAs
  filter(count_group == 'count ≥ 5') %>%
  # to use all val together
  filter(!is.na(amp_seq_val), !is.na(RR_val)) %>%
  mutate(all_val = ifelse(qPCR_val == RR_val & qPCR_val == amp_seq_val & qPCR_val == 'pass', 'pass', 'fail')) %>%
  # change nr of databases to binary
  mutate(n_db = ifelse(is.na(n_db), 0, 1)) %>%
  group_by(n_db) %>%
  count(all_val) %>%
  pivot_wider(values_from = n, names_from = all_val) %>%
  ungroup() %>%
  select(-n_db)


cont_table = as.data.frame(cont_table)
rownames(cont_table) <- c("novel", "in_db")

cont_table

plot data

mosaicplot(cont_table,
           main = "Mosaic plot",
           color = TRUE)

if chisquare expected < 5 => then Fisher should be used

chisq.test(cont_table)$expected

##           fail     pass
## novel 16.23933 132.7607
## in_db 80.76067 660.2393

chisquare test

chisq.test(cont_table)

## 
##  Pearson's Chi-squared test with Yates' continuity correction
## 
## data:  cont_table
## X-squared = 162.61, df = 1, p-value < 2.2e-16

OR = (705/36)/(88/61)
OR

## [1] 13.57481

Link with detected by multiple tools (n = 890)

make count table

cont_table = cq %>%
  select(circ_id, cell_line, qPCR_val, RR_val, amp_seq_val, count_group, n_detected) %>%
  unique() %>%
  filter(count_group == 'count ≥ 5') %>%
  # to use all val together
  filter(!is.na(amp_seq_val), !is.na(RR_val)) %>%
  mutate(all_val = ifelse(qPCR_val == RR_val & qPCR_val == amp_seq_val & qPCR_val == 'pass', 'pass', 'fail')) %>%
  # change nr of databases to binary
  mutate(n_detected = ifelse(n_detected == 1, 0, 1)) %>%
  group_by(n_detected) %>%
  count(all_val) %>%
  pivot_wider(values_from = n, names_from = all_val) %>%
  ungroup() %>%
  select(-n_detected)


cont_table = as.data.frame(cont_table)
rownames(cont_table) <- c("unique", "multiple_tools")

cont_table

plot data

mosaicplot(cont_table,
           main = "Mosaic plot",
           color = TRUE)

if chisquare expected < 5 => then Fisher should be used

chisq.test(cont_table)$expected

##                     fail     pass
## unique          7.738202  63.2618
## multiple_tools 89.261798 729.7382

chisquare test

chisq.test(cont_table)

## 
##  Pearson's Chi-squared test with Yates' continuity correction
## 
## data:  cont_table
## X-squared = 330.06, df = 1, p-value < 2.2e-16

OR = (776/43)/(17/54)
OR

## [1] 57.32421

Link with nr of exons

Single exon vs multi-exon (n = 723)

make count table

cont_table = cq %>%
  select(circ_id, cell_line, qPCR_val, RR_val, amp_seq_val, count_group, nr_exons) %>%
  unique() %>%
  filter(count_group == 'count ≥ 5') %>%
  # to use all val together
  filter(!is.na(amp_seq_val), !is.na(RR_val)) %>%
  mutate(all_val = ifelse(qPCR_val == RR_val & qPCR_val == amp_seq_val & qPCR_val == 'pass', 'pass', 'fail')) %>%
  # also remove the ones that do not have a exon annotation
  filter(!is.na(nr_exons), !nr_exons == "ambiguous") %>%
  mutate(exon_bin = ifelse(nr_exons == 1, 0, 1)) %>%
  group_by(exon_bin) %>%
  count(all_val) %>% 
  pivot_wider(values_from = n, names_from = all_val) %>%
  ungroup() %>%
  mutate(fail = ifelse(is.na(fail), 0, fail)) %>%
  select(-exon_bin)

cont_table = as.data.frame(cont_table)
rownames(cont_table) <- c("single_exon", "multi_exons")
cont_table

plot data

mosaicplot(cont_table,
           main = "Mosaic plot",
           color = TRUE)

if chisquare expected < 5 => then Fisher should be used

chisq.test(cont_table)$expected

##                 fail     pass
## single_exon  7.69018 131.3098
## multi_exons 32.30982 551.6902

chisquare test

chisq.test(cont_table)

## 
##  Pearson's Chi-squared test with Yates' continuity correction
## 
## data:  cont_table
## X-squared = 16.399, df = 1, p-value = 5.13e-05

OR = (562/22)/(121/18)
OR

## [1] 3.80015

Is ‘exon 1’ included in the circRNA? (n = 592)

cq_start = cq %>% 
  mutate(start_exon_nr = substr(start_match, 19, 27),
         start_exon_nr = ifelse(substr(start_exon_nr, 1, 1) == '_',
                                substr(start_exon_nr, 2, 10),
                                start_exon_nr))
cq_start

cont_table = cq_start %>%
  filter(count_group == 'count ≥ 5') %>%
  select(circ_id, cell_line, qPCR_val, RR_val, amp_seq_val, start_exon_nr) %>%
  unique() %>%
  filter(!is.na(start_exon_nr)) %>%
  mutate(exon_1 = ifelse(start_exon_nr == "exon_1", 1, 0)) %>%
  # to use all val together
  filter(!is.na(amp_seq_val), !is.na(RR_val)) %>%
  mutate(all_val = ifelse(qPCR_val == RR_val & qPCR_val == amp_seq_val & qPCR_val == 'pass', 
                          'pass', 'fail')) %>%
  # change nr of databases to binary
  group_by(exon_1) %>%
  count(all_val) %>%
  pivot_wider(values_from = n, names_from = all_val) %>%
  ungroup() 


cont_table

=> not enough values

Canonical splicing (n = 722)

cont_table = cq %>%
  filter(count_group == 'count ≥ 5',
         !strand == 'unknown') %>%
  select(circ_id, cell_line, qPCR_val, RR_val, amp_seq_val, ss_motif) %>%
  unique() %>%
  mutate(ss_can = ifelse(ss_motif == "AGNGT", 1, 0)) %>%
  # to use all val together
  filter(!is.na(amp_seq_val), !is.na(RR_val)) %>%
  mutate(all_val = ifelse(qPCR_val == RR_val & qPCR_val == amp_seq_val & qPCR_val == 'pass', 
                          'pass', 'fail')) %>%
  # change nr of databases to binary
  group_by(ss_can) %>%
  count(all_val) %>%
  pivot_wider(values_from = n, names_from = all_val) %>%
  ungroup() %>%
  select(-ss_can)


cont_table = as.data.frame(cont_table)
rownames(cont_table) <- c("non_cannonical", "cannonical")

cont_table

plot data

mosaicplot(cont_table,
           main = "Mosaic plot",
           color = TRUE)

if chisquare expected < 5 => then Fisher should be used

chisq.test(cont_table)$expected

##                    fail      pass
## non_cannonical 11.99169  99.00831
## cannonical     66.00831 544.99169

chisquare test

chisq.test(cont_table)

## 
##  Pearson's Chi-squared test with Yates' continuity correction
## 
## data:  cont_table
## X-squared = 42.044, df = 1, p-value = 8.923e-11

OR = (565/46)/(79/32)
OR

## [1] 4.975234

Known annotation (n = 891)

cont_table = cq %>%
  filter(count_group == "count ≥ 5") %>%
  select(circ_id, cell_line, qPCR_val, RR_val, amp_seq_val, ENST) %>%
  unique() %>%
  # to use all val together
  filter(!is.na(amp_seq_val), !is.na(RR_val)) %>%
  mutate(all_val = ifelse(qPCR_val == RR_val & qPCR_val == amp_seq_val & qPCR_val == 'pass', 
                          'pass', 'fail'),
         ENST_group = ifelse(is.na(ENST), 'no_match', "match")) %>% 
  # change nr of databases to binary
  group_by(ENST_group) %>%
  count(all_val) %>%
  pivot_wider(values_from = n, names_from = all_val) %>%
  ungroup() %>%
  select(-ENST_group)


cont_table = as.data.frame(cont_table)
rownames(cont_table) <- c("match", "no_match")

cont_table

plot data

mosaicplot(cont_table,
           main = "Mosaic plot",
           color = TRUE)

if chisquare expected < 5 => then Fisher should be used

chisq.test(cont_table)$expected

##               fail      pass
## match    87.637486 717.36251
## no_match  9.362514  76.63749

chisquare test

chisq.test(cont_table)

## 
##  Pearson's Chi-squared test with Yates' continuity correction
## 
## data:  cont_table
## X-squared = 203.2, df = 1, p-value < 2.2e-16

OR = (757/48)/(37/49)
OR

## [1] 20.8857

CircRNA detection tool approach (n = 798)

cont_table = cq %>%
  filter(count_group == "count ≥ 5") %>%
  left_join(read_tsv('../data/details/tool_annotation.txt')) %>%
  select(circ_id, cell_line, qPCR_val, RR_val, amp_seq_val, approach) %>%
  filter(!approach == 'integrative') %>%
  unique() %>%
  # to use all val together
  filter(!is.na(amp_seq_val), !is.na(RR_val)) %>%
  mutate(all_val = ifelse(qPCR_val == RR_val & qPCR_val == amp_seq_val & qPCR_val == 'pass', 
                          'pass', 'fail')) %>% 
  # change nr of databases to binary
  group_by(approach) %>%
  count(all_val) %>%
  pivot_wider(values_from = n, names_from = all_val) %>%
  ungroup() %>%
  select(-approach)

## Rows: 16 Columns: 7
## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: "\t"
## chr (7): tool, approach, tool_lt, lin_annotation, strand_anno, splicing, BSJ...
## 
## ℹ Use `spec()` to retrieve the full column specification for this data.
## ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.
## Joining, by = "tool"

cont_table = as.data.frame(cont_table)
rownames(cont_table) <- c("candidate-based", "segmented read-based")

cont_table

plot data

mosaicplot(cont_table,
           main = "Mosaic plot",
           color = TRUE)

if chisquare expected < 5 => then Fisher should be used

chisq.test(cont_table)$expected

##                          fail     pass
## candidate-based      21.54762 159.4524
## segmented read-based 73.45238 543.5476

chisquare test

chisq.test(cont_table)

## 
##  Pearson's Chi-squared test with Yates' continuity correction
## 
## data:  cont_table
## X-squared = 6.8785, df = 1, p-value = 0.008724

OR = (170/11)/(533/84)
OR

## [1] 2.435613

BSJ count group (n = 1042)

cont_table = cq %>%
  filter(!count_group == "no_counts") %>%
  select(circ_id, cell_line, qPCR_val, RR_val, amp_seq_val, count_group) %>%
  unique() %>%
  # to use all val together
  filter(!is.na(amp_seq_val), !is.na(RR_val)) %>%
  mutate(all_val = ifelse(qPCR_val == RR_val & qPCR_val == amp_seq_val & qPCR_val == 'pass', 
                          'pass', 'fail')) %>%
  # change nr of databases to binary
  group_by(count_group) %>%
  count(all_val) %>%
  pivot_wider(values_from = n, names_from = all_val) %>%
  ungroup() %>%
  select(-count_group)


cont_table = as.data.frame(cont_table)
rownames(cont_table) <- c("count < 5", "count ≥ 5")

cont_table

plot data

mosaicplot(cont_table,
           main = "Mosaic plot",
           color = TRUE)

if chisquare expected < 5 => then Fisher should be used

chisq.test(cont_table)$expected

##                fail     pass
## count < 5  19.83877 132.1612
## count ≥ 5 116.16123 773.8388

chisquare test

chisq.test(cont_table)

## 
##  Pearson's Chi-squared test with Yates' continuity correction
## 
## data:  cont_table
## X-squared = 23.636, df = 1, p-value = 1.164e-06

OR = (793/97)/(113/39)
OR

## [1] 2.821549

Using sensitivity

tool_anno = read_tsv('../data/details/tool_annotation.txt')

## Rows: 16 Columns: 7
## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: "\t"
## chr (7): tool, approach, tool_lt, lin_annotation, strand_anno, splicing, BSJ...
## 
## ℹ Use `spec()` to retrieve the full column specification for this data.
## ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

val = read_tsv('../data/Supplementary_Table_4_precision_values.txt')

## Rows: 29 Columns: 16
## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: "\t"
## chr  (2): tool, count_group
## dbl (14): nr_qPCR_total, nr_qPCR_fail, nr_qPCR_val, nr_RR_total, nr_RR_fail,...
## 
## ℹ Use `spec()` to retrieve the full column specification for this data.
## ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

add annotation to sensitivity

sens_anno = val %>% 
  group_by(tool) %>%
  slice(1) %>%
  select(tool, sensitivity) %>%
  left_join(tool_anno) %>%
  select(-tool_lt) %>% ungroup()

## Joining, by = "tool"

sens_anno

Detection approach: segmented read-based VS candidate-based

wilcox.test(sensitivity ~ approach, data=sens_anno %>% filter(!approach == 'integrative'))

## 
##  Wilcoxon rank sum exact test
## 
## data:  sensitivity by approach
## W = 6, p-value = 0.1264
## alternative hypothesis: true location shift is not equal to 0

kruskal.test(sensitivity ~ approach, data=sens_anno)

## 
##  Kruskal-Wallis rank sum test
## 
## data:  sensitivity by approach
## Kruskal-Wallis chi-squared = 3.3209, df = 2, p-value = 0.1901

Detection based on annotation: known splice sites VS entire genome

wilcox.test(sensitivity ~ lin_annotation, data=sens_anno)

## 
##  Wilcoxon rank sum exact test
## 
## data:  sensitivity by lin_annotation
## W = 22, p-value = 0.8
## alternative hypothesis: true location shift is not equal to 0

Strand annotation method

kruskal.test(sensitivity ~ strand_anno, data=sens_anno %>% 
               filter(!strand_anno == "no strand reported"))

## 
##  Kruskal-Wallis rank sum test
## 
## data:  sensitivity by strand_anno
## Kruskal-Wallis chi-squared = 5.2538, df = 3, p-value = 0.1541

Splicing: canonical VS non-canonical

wilcox_ss = wilcox.test(sensitivity ~ splicing, data=sens_anno) 

wilcox_ss

## 
##  Wilcoxon rank sum exact test
## 
## data:  sensitivity by splicing
## W = 55, p-value = 0.0004579
## alternative hypothesis: true location shift is not equal to 0

sens_anno %>%
  group_by(splicing) %>%
  summarise(mean_sens = mean(sensitivity))

method 1

N = nrow(sens_anno)
N

## [1] 16

Z = qnorm(wilcox_ss$p.value/2)
Z

## [1] -3.504262

r = abs(Z)/sqrt(N)
r

## [1] 0.8760654

method 2

library(rstatix)
library(ggpubr)
sens_anno %>% rstatix::wilcox_test(sensitivity ~ splicing)

sens_anno %>% rstatix::wilcox_effsize(sensitivity ~ splicing)

#sens_anno %>% rstatix::wilcox_effsize(sensitivity ~ splicing, ci = T)

Median difference in sensitivity

median_diff = tibble()

for (tool_can in sens_anno %>% filter(splicing == "canonical") %>% pull(tool)) {
  sens_can = sens_anno %>% filter(tool == tool_can) %>% pull(sensitivity)
  
  for (tool_non_can in sens_anno %>% filter(splicing == "non-canonical") %>% pull(tool)) {
    sens_non_can = sens_anno %>% filter(tool == tool_non_can) %>% pull(sensitivity)
    
    median_diff = median_diff %>%
      bind_rows(tibble(tool_can, sens_can, tool_non_can, sens_non_can))
    
  }
}

median_diff = median_diff %>%
  mutate(sens_diff = sens_can - sens_non_can)

median_diff

median_diff %>% pull(sens_diff) %>% median()

## [1] 0.384535

Filtering

wilcox.test(sensitivity ~ BSJ_filter, data=sens_anno)

## 
##  Wilcoxon rank sum exact test
## 
## data:  sensitivity by BSJ_filter
## W = 35, p-value = 0.4409
## alternative hypothesis: true location shift is not equal to 0

Link between circRNA annotation and validation rates

Marieke Vromman

2022-12-05

File set-up

Set working directory to current directory

Load standard libraries and resolve conflicts

Read data

Presence in db (n = 890)

Link with detected by multiple tools (n = 890)

Link with nr of exons

Single exon vs multi-exon (n = 723)

Is ‘exon 1’ included in the circRNA? (n = 592)

Canonical splicing (n = 722)

Known annotation (n = 891)

CircRNA detection tool approach (n = 798)

BSJ count group (n = 1042)

Using sensitivity

Detection approach: segmented read-based VS candidate-based

Detection based on annotation: known splice sites VS entire genome

Strand annotation method

Splicing: canonical VS non-canonical

method 1

method 2

Median difference in sensitivity

Filtering